I've kept the actual FR10 cover samples, not just the pictures, but now one accepted cover gets a bright line every time it sits slightly differently in the holder. Our genuine scratched cover barely shows its mark in that view. Feels backwards, tbh.
Jonas, the rocking needs attention, but Mia also says the real scratch is poorly visible. A corrected holder could produce consistently inadequate images. Keep that scratched cover in the comparison.
Harish is right. I once spent a morning getting an inspection image beautifully repeatable, then the inspector brought over the reject we were supposed to find. My new support hid its damaged edge.
Mia, ask the tooling person to look at the actual seating and ask quality to keep both covers identified. After the holder correction, compare the accepted cover and the real scratch under each candidate lighting arrangement. Don't throw away the awkward scratch image just because a brighter picture of the good cover looks more persuasive.
And check the corrected holder leaves the whole required surface visible; our narrow edge was easy to lose behind a support even with the part seated properly
Tooling found a pad from the thicker cover family. Correct pad fitted, quality can reseat both samples without that rock now. Thanks David, keeping the real reject there stopped us celebrating the first tidy picture.
Only one presentation so far. We're keeping the holder correction, but the camera person has taken the lighting comparison back. And no, the threshold hasn't magically become sensible.
Then label those images as a comparison in progress. Otherwise the nicely seated good cover will find its way into a finished-example folder before the scratched one catches up.